A machine-vision framework for electric vehicle charging hole localization based on contour-driven detection and geometric verification
- 1Nanjing University of Science and Technology Zijin College, Nanjing, China.
- 2NARI Technology Development Co. Ltd, Nanjing, China.
- 3State Grid Electric Power Research Institute, Nanjing, China.
Abstract
The accurate localization of charging holes is an indispensable perception step for the automatic charging system of electric vehicles, since the success of subsequent charging-gun insertion depends directly on the precise estimation of the geometric centers of the holes on the vehicle charging interface. This paper presents a lightweight machine-vision framework for charging hole localization based on contour-driven detection and geometric verification. The proposed framework integrates image filtering, RGB-to-HSL color-space conversion, parameterized HSL adjustment, two-stage contour detection, minimum enclosing circle fitting, and Hough circle verification into a deterministic pipeline. The framework is interpretable, does not require a learning-based training process, and can be deployed on conventional CPU-based computing resources. The image data used in this study were acquired by a monocular RGB camera mounted on the charging robot end-effector, in the same sensing configuration in which the system is intended to operate. A total of 80 images were collected from six vehicles equipped with the same five-hole charging interface, under realistic variations of illumination, viewing angle, reflective interference, and background complexity. Detailed parameter selection rationale, a parameter perturbation analysis, and a condition-specific ablation analysis are reported. On an Intel Core i7-12700 CPU with 16 GB RAM and no GPU acceleration, the proposed framework achieves an overall mean localization error of 4.2 pixels and a 96.25% detection success rate, with an average per-image processing time of approximately 60 ms. The results demonstrate that the proposed framework provides accurate, robust, and engineering-deployable charging-hole localization for the target charging interface.
Related
- A real-time deep learning method for detection and localization of non-emerged cabbage seedlings
- A method for brain tumor detection and localization by means of deep learning and reliable prediction explainability
- The interplay between detection and localization in human vision
- Lightweight YOLOv8n marigold corolla detection and picking-point localization method based on ECR progressive channel compression and dual attention mechanisms
- Morphology-guided dual-level detection and depth localization of cotton terminal buds for precision topping
BibTeX
@article{luo2026machinevision,
title = {{A machine-vision framework for electric vehicle charging hole localization based on contour-driven detection and geometric verification}},
author = {Wei Luo and Qiantao Huo and Xin Li},
journal = {PLOS One},
year = {2026},
month = oct,
doi = {10.1371/journal.pone.0359570},
url = {https://doi.org/10.1371/journal.pone.0359570},
note = {CC BY},
}
RIS
TY - JOUR
TI - A machine-vision framework for electric vehicle charging hole localization based on contour-driven detection and geometric verification
AU - Wei Luo
AU - Qiantao Huo
AU - Xin Li
PY - 2026
DA - 2026/10/22
JO - PLOS One
DO - 10.1371/journal.pone.0359570
UR - https://doi.org/10.1371/journal.pone.0359570
AB - The accurate localization of charging holes is an indispensable perception step for the automatic charging system of electric vehicles, since the success of subsequent charging-gun insertion depends directly on the precise estimation of the geometric centers of the holes on the vehicle charging interface. This paper presents a lightweight machine-vision framework for charging hole localization based on contour-driven detection and geometric verification. The proposed framework integrates image filtering, RGB-to-HSL color-space conversion, parameterized HSL adjustment, two-stage contour detection, minimum enclosing circle fitting, and Hough circle verification into a deterministic pipeline. The framework is interpretable, does not require a learning-based training process, and can be deployed on conventional CPU-based computing resources. The image data used in this study were acquired by a monocular RGB camera mounted on the charging robot end-effector, in the same sensing configuration in which the system is intended to operate. A total of 80 images were collected from six vehicles equipped with the same five-hole charging interface, under realistic variations of illumination, viewing angle, reflective interference, and background complexity. Detailed parameter selection rationale, a parameter perturbation analysis, and a condition-specific ablation analysis are reported. On an Intel Core i7-12700 CPU with 16 GB RAM and no GPU acceleration, the proposed framework achieves an overall mean localization error of 4.2 pixels and a 96.25% detection success rate, with an average per-image processing time of approximately 60 ms. The results demonstrate that the proposed framework provides accurate, robust, and engineering-deployable charging-hole localization for the target charging interface.
SN - 1932-6203
ER -